1. Field of the Invention
The invention relates to database pointer error diagnostics. Particularly, the invention relates to autonomic Information Management System (IMS) mainframe database pointer error diagnostic data extraction.
2. Description of the Related Art
When IMS databases on mainframe computers experience a “pointer error” showing that the database is damaged in some way, the database administrator will often collect diagnostic data in an effort to identify the problem. However, this problem source identification and diagnostic data collection is fairly complicated and require significant experience and specific skill. Furthermore, the large quantities of IMS log and trace data requiring analysis makes the problem source identification of a database pointer error a very time consuming operation.
Application abends, messages, ad-hoc utilities output, and pointer checker reports can provide diagnostic clues identifying the affected database name, the associated pointer errors, and other relevant information. However, the process of determining the cause of a database pointer error includes manually locating a clean Image Copy (IC) of the database for recovery usage and running a pointer checker utility on each potential IC to determine if the IC is suitable for recovery.
In addition, sifting through the enormous amount of log records between the error occurrence and the pointer error detection is a daunting task involving the screening of millions of records.